Arabic digits

John C Klensin klensin at jck.com
Thu Dec 4 14:53:40 CET 2008



--On Thursday, 04 December, 2008 17:55 +0900 Martin Duerst
<duerst at it.aoyama.ac.jp> wrote:

> Short summary: Mostly correct. My conclusion (already expressed
> in an earlier mail today) is that the best way to deal with the
> equivalence problem is to do automatic parallel registration,
> without affecting the protocol.

Martin,

I think (I'm still a little confused and trying to keep an open
mind) that I'd agree if we could trust all registries / zone
administrators at all levels to do that parallel registration
and behave themselves.  But we know there are rogue
registrations out there and, more important, banning
carelessness is not a productive activity.

Also, while I think it unlikely that one would want to do it in
an environment in which two people hit the same key on
seemingly-identical keyboards and get two different coded
characters, I can imagine situations in which a registry
actually would want to treat the two as different.

Those concerns, especially the former, lead me to believe that
imposing a protocol restriction to ensure homogeneous digits in
combination with Arabic script is probably a good idea, giving
us a slightly better handle on having the right thing happen and
encouraging _all_ relevant registries to cope with the issue and
make decisions about it.   At least at the moment, I think the
tradeoffs balance out that way, with the understanding that (i)
nothing we could do in the protocol will "solve" the problem (it
is just a matter of enabling, supporting, and encouraging
registry decisions and actions, including the parallel
registrations that you suggest) and (ii) there is enough risk
here (because of the user's inability to control what is
actually coded, not because of phishing or visual confusion) to
justify our taking an extra step to strengthen things even if
registries are a sloppy.

YMMD, of course (and I may change my mind after thinking more
about it).
    john





More information about the Idna-update mailing list